Picture for Ryan Feng

Ryan Feng

Achieving the Safety and Security of the End-to-End AV Pipeline

Add code
Sep 05, 2024
Viaarxiv icon

Theoretically Principled Trade-off for Stateful Defenses against Query-Based Black-Box Attacks

Add code
Jul 30, 2023
Viaarxiv icon

Investigating Stateful Defenses Against Black-Box Adversarial Examples

Add code
Mar 17, 2023
Viaarxiv icon

Constraining the Attack Space of Machine Learning Models with Distribution Clamping Preprocessing

Add code
May 18, 2022
Figure 1 for Constraining the Attack Space of Machine Learning Models with Distribution Clamping Preprocessing
Figure 2 for Constraining the Attack Space of Machine Learning Models with Distribution Clamping Preprocessing
Figure 3 for Constraining the Attack Space of Machine Learning Models with Distribution Clamping Preprocessing
Figure 4 for Constraining the Attack Space of Machine Learning Models with Distribution Clamping Preprocessing
Viaarxiv icon

Concept-based Explanations for Out-Of-Distribution Detectors

Add code
Mar 04, 2022
Figure 1 for Concept-based Explanations for Out-Of-Distribution Detectors
Figure 2 for Concept-based Explanations for Out-Of-Distribution Detectors
Figure 3 for Concept-based Explanations for Out-Of-Distribution Detectors
Figure 4 for Concept-based Explanations for Out-Of-Distribution Detectors
Viaarxiv icon

Towards Adversarially Robust Deepfake Detection: An Ensemble Approach

Add code
Feb 11, 2022
Figure 1 for Towards Adversarially Robust Deepfake Detection: An Ensemble Approach
Figure 2 for Towards Adversarially Robust Deepfake Detection: An Ensemble Approach
Figure 3 for Towards Adversarially Robust Deepfake Detection: An Ensemble Approach
Figure 4 for Towards Adversarially Robust Deepfake Detection: An Ensemble Approach
Viaarxiv icon

Using Anomaly Feature Vectors for Detecting, Classifying and Warning of Outlier Adversarial Examples

Add code
Jul 01, 2021
Figure 1 for Using Anomaly Feature Vectors for Detecting, Classifying and Warning of Outlier Adversarial Examples
Figure 2 for Using Anomaly Feature Vectors for Detecting, Classifying and Warning of Outlier Adversarial Examples
Figure 3 for Using Anomaly Feature Vectors for Detecting, Classifying and Warning of Outlier Adversarial Examples
Figure 4 for Using Anomaly Feature Vectors for Detecting, Classifying and Warning of Outlier Adversarial Examples
Viaarxiv icon

Smart Black Box 2.0: Efficient High-bandwidth Driving Data Collection based on Video Anomalies

Add code
Feb 09, 2021
Figure 1 for Smart Black Box 2.0: Efficient High-bandwidth Driving Data Collection based on Video Anomalies
Figure 2 for Smart Black Box 2.0: Efficient High-bandwidth Driving Data Collection based on Video Anomalies
Figure 3 for Smart Black Box 2.0: Efficient High-bandwidth Driving Data Collection based on Video Anomalies
Figure 4 for Smart Black Box 2.0: Efficient High-bandwidth Driving Data Collection based on Video Anomalies
Viaarxiv icon

Essential Features: Reducing the Attack Surface of Adversarial Perturbations with Robust Content-Aware Image Preprocessing

Add code
Dec 03, 2020
Figure 1 for Essential Features: Reducing the Attack Surface of Adversarial Perturbations with Robust Content-Aware Image Preprocessing
Figure 2 for Essential Features: Reducing the Attack Surface of Adversarial Perturbations with Robust Content-Aware Image Preprocessing
Figure 3 for Essential Features: Reducing the Attack Surface of Adversarial Perturbations with Robust Content-Aware Image Preprocessing
Figure 4 for Essential Features: Reducing the Attack Surface of Adversarial Perturbations with Robust Content-Aware Image Preprocessing
Viaarxiv icon

Query-Efficient Physical Hard-Label Attacks on Deep Learning Visual Classification

Add code
Feb 17, 2020
Figure 1 for Query-Efficient Physical Hard-Label Attacks on Deep Learning Visual Classification
Figure 2 for Query-Efficient Physical Hard-Label Attacks on Deep Learning Visual Classification
Figure 3 for Query-Efficient Physical Hard-Label Attacks on Deep Learning Visual Classification
Figure 4 for Query-Efficient Physical Hard-Label Attacks on Deep Learning Visual Classification
Viaarxiv icon